lymph node metastasis in early gastric cancer · lymph node metastasis in early gastric cancer. of...

Bravo NetoBravo NetoBravo NetoBravo NetoBravo NetoLymph node metastasis in early gastric cancer 11

Rev. Col. Bras. Cir. 2014; 41(1): 011-017

Original ArticleOriginal ArticleOriginal ArticleOriginal ArticleOriginal Article

Lymph node metastasis in early gastric cancerLymph node metastasis in early gastric cancerLymph node metastasis in early gastric cancerLymph node metastasis in early gastric cancerLymph node metastasis in early gastric cancer

Metástase linfonodal em câncer gástrico precoceMetástase linfonodal em câncer gástrico precoceMetástase linfonodal em câncer gástrico precoceMetástase linfonodal em câncer gástrico precoceMetástase linfonodal em câncer gástrico precoce

GUILHERME PINTO BRAVO NETO, TCBC-RJ1; ELIZABETH GOMES DOS SANTOS,TCBC-RJ 2; FELIPE CARVALHO VICTER, TCBC-RJ2;CARLOS EDUARDO DE SOUZA CARVALHO3

A B S T R A C TA B S T R A C TA B S T R A C TA B S T R A C TA B S T R A C T

ObjectiveObjectiveObjectiveObjectiveObjective: to evaluate the incidence of lymph node metastasis in early gastric cancer, identifying risk factors for its

development. MethodsMethodsMethodsMethodsMethods: we conducted a prospective study of patients with gastric cancer admitted to the Section of the

Esophago-Gastric Surgery of the Surgery of Service HUCFF-UFRJ, from January 2006 to May 2012. ResultsResultsResultsResultsResults: the rate of

early gastric cancer was 16.3%. The incidence of nodal metastases was 30.8% and occurred more frequently in patients

with tumors with involvement of the submucosa (42.9%), in those poorly differentiated (36.4%), in tumors larger than 2 cm

(33.3%) and in type III ulcerated lesions (43.8%). ConclusionConclusionConclusionConclusionConclusion: the incidence of lymph node metastases in patients was very

high and suggests that one should keep the radicality of resection in early gastric cancer, particularly in relation to D2

lymphadenectomy, recommended for advanced gastric cancer. Conservative resections, with lymphadenectomies smaller

than D2, should be performed only in selected cases, well-studied as for the risk factors of lymph node metastasis. Despite

the small number of cases did not permit to relate the rate of lymph node metastasis to the risk factors considered, we

noted a strong tendency for the occurrence of these metastases in the poorly differentiated, type III, larger than 2 cm

tumors, and in the Lauren diffuse types.

Key words:Key words:Key words:Key words:Key words: Stomach neoplasms. Lymphatic metastasis. Gastrectomy. Lymph node excision. Risk factors.

1. Departament of Surgery, Faculty of Medicine, UFRJ; 2. Esophago-Gastric Section, General Surgery Service, Clementino Fraga Filho UniversitaryHospital, UFRJ; 3. Service of Pathology, Clementino Fraga Filho Universitary Hospital, UFRJ.

INTRODUCTIONINTRODUCTIONINTRODUCTIONINTRODUCTIONINTRODUCTION

Lymph node metastasis in gastric cancer is frequent andearly, and depends on many variables, amongst which

the depth of invasion of the gastric wall, the degree oftumor differentiation and tumor size1-4. Thus, early tumorsmay have incidence of lymph node metastases rangingfrom 0%, when restricted to the mucosa, welldifferentiated and less than 2cm in size, to more than30%, when reaching the submucosa, being poorlydifferentiated and having more than 2cm diameter5,6. Theknowledge of the parameters of risk for these metastases,as well as the use of diagnostic procedures able to identifythem in the preoperative and intraoperative periods, madepossible the realization of more conservative procedures,such as endoscopic resection and smaller gastrectomieswith limited lymphadenectomy, in selected cases of earlygastric cancer, thus individualizing surgery of gastricadenocarcinoma and reducing treatment complications,both early and late7,8.

The objective of this study was to evaluate theincidence of lymph node metastasis in early gastric cancer(EGC), identifying its risk factors for its development.

METHODSMETHODSMETHODSMETHODSMETHODS

We conducted a prospective study of patientswith gastric cancer admitted to the Esophago-Gastric Sectionof the General Surgery Service of the Clementino FragaFilho University Hospital, Federal University of Rio de Janei-ro (HUCFF–UFRJ), from January 2006 to May 2012. Allpatients admitted with a diagnosis of gastricadenocarcinoma confirmed by histopathologicalexamination of gastric tissue obtained by endoscopicbiopsies fragment underwent clinical staging and imaging.Gastric cancer patients without criteria of inoperability orunresectability, confirmed by laparoscopy or laparotomy,were submitted to curative gastric resection with D2lymphadenectomy. All patients were informed of the risksand benefits of the procedure and signed an informedconsent form.

From 2008 on, with the introduction of endoscopicultrasonography in HUCFF–UFRJ, patients with early gastriclesions suggestive of cancer at endoscopy underwentendoscopic ultrasound to confirm the degree of tumorpenetration of the gastric wall and evaluation of suspectperigastric nodes. Those with echoendoscopic confirmation

Gisele Higa

Texto digitado

10.1590/S0100-69912014000100004

Gisele Higa

Texto digitado

12

Rev. Col. Bras. Cir. 2014; 41(1): 011-017

Bravo NetoBravo NetoBravo NetoBravo NetoBravo NetoLymph node metastasis in early gastric cancer

of early tumor without perigastric lymphadenopathy wereoperated and submitted to sentinel nodes research, which,when negative, provided the realization of smaller gastricresections with modified D1 lymphadenectomy, asrecommended by the Japanese Society of Gastric Cancer9,10.Patients with endoscopic ultrasound doubtful as to thedegree of penetration of the tumor and/or suspect perigastriclymph nodes were treated for advanced tumors. For all thepatients operated the lymphadenectomy specimen wasindividualized as for the number of resected lymph nodesand the presence of nodal metastases.

The macroscopic appearance of the lesion wasclassified according to criteria of the Japanese Society ofGastric Cancer11 in types I, IIa, IIb, IIc and III. Itsdimensions and microscopic features, such as cellulardifferentiation and Lauren classification, were evaluatedin surgical specimens. The parameters used for the riskof nodal metastases configuration were: tumor size, lessthan or equal to 2 cm versus larger than 2 cm;macroscopic classification: type I and II versu type III;degree of cell differentiation: well- or moderatelydifferentiated versus poorly differentiated; Laurenclassification: intestinal type versus diffuse type; degreeof tumor penetration of the gastric wall: restricted to themucosa (T1a) versus with involvement of the submucosa(T1b).

The staging of gastric cancer followed the criteriaestablished by the American Joint Committee on Cancer(AJCC) in its seventh edition, 201012.

The descriptive analysis of the observed data waspresented in the form of tables, expressed by the frequency(n) and percentage (%) for categorical data and by median,minimum and maximum for numeric data. In order to checkwhether there was a significant association between thevariables of the lesion with the presence of metastasis, weapplied: the chi-square (2) or Fisher exact test forcomparisons of categorical (qualitative) data and the Mann-Whitney test (nonparametric) to compare numerical data(lesion size in cm).

We applide the nonparametric method, becausethe size of the lesion did not show normal distribution(Gaussian distribution) due to the wide dispersion of thedata and rejection of the hypothesis of normality accordingto the Kolmogorov-Smirnov test. The criterion fordetermining significance was set at 5%.

RESULTSRESULTSRESULTSRESULTSRESULTS

During the study period, 160 patients withgastric cancer were admitted to staging. Of these, nine(5.6%) were considered inoperable, 12 (7.5%)unresectable at laparotomy, and 14 (8.8%) weresubmitted to palliative gastrojejunostomy due to thepresence of pyloric obstruction. The remaining 125(78.1%) underwent gastric resections, of which 83(66.4%) with curat ive intent, with D2lymphadenectomy, 30 (24%) with palliative resectionsfor locally advanced disease without possibility of R0resection, and 12 (9.6%) with atypical resections andgastric D1 modified lymphadenectomy (perigastric nodesplus 7 and/or 8 and/or 9 chains) (Table 1). Among thelatter, 11 were patients with early tumors withoutevidence of lymph node metastasis at sentinel nodesresearch and one had tumor in situ.

The mean number of resected lymph nodesin D2 and modified D1 lymphadenectomies was 25. Ofthe patients who underwent gastrectomy with D2lymphadenectomy, D1 (palliative resections) andmodified D1, 116 could be properly staged. Theincidence of lymph node metastases (N) and stagingbased on the degree of penetration of the tumor in thegastric wall (T) can be seen in Table 2. There was asignificant association (p<0.0001) between the degreeof penetration of the tumor in the gastric wall (T) andlymph node metastasis (N +). The list of patients withEGC and the tumors features, such as the degree ofpenetration of the gastric wall (T1a – restricted to themucosa – or T1b – with submucosal involvement), thesize of the lesion, macroscopic classification and Lauren,the degree of tumor differentiation and the presenceof lymph node metastases are shown in Table 3.

The frequency of EGC was 16.3% (26patients), 7.5% restricted to the mucosa (12 patients)and 8.8% with involvement of the submucosa (14patients). The incidence of nodal metastases in patientswith EGC was 30.8% and occurred more frequently inpatients with tumors with involvement of the submucosa(42.9%), those moderately/poorly differentiated(38.9%), greater than 2cm (33.3%) and ulcerated typeIII (43.8%) (Tables 4 and 5). The relationship betweenthe rates of lymph node metastasis and the variables of

Table 1 -Table 1 -Table 1 -Table 1 -Table 1 - Approach to patients without preoperative inoperability criteria.

Operable PatientsOperable PatientsOperable PatientsOperable PatientsOperable Patients N (%)N (%)N (%)N (%)N (%) Surgery PerformedSurgery PerformedSurgery PerformedSurgery PerformedSurgery Performed

Local advanced disease without possibility of 30 (24%) Palliative Resection

R0 resection

Operable with possibility of R0 resection 83 (66.4%) Gastrectomy with D2 lymphadenectomy

Early tumors without lymph node metastasis 12 (9,6%) Atypical Gastrectomy with limited

lymphadenectomy


Rev. Col. Bras. Cir. 2014; 41(1): 011-017

the lesion, such as the degree of penetration of thegastric wall (T), lesion size, macroscopic classification,tumor differentiation grade and histological type ofLauren are shown in Table 6. There was no significantassociation between the variables of the lesion withthe presence of metastases due to the small number ofcases studied. There was, however, a trend toward anincreased risk of metastasis in tumors larger than 2cm,poorly differentiated, diffuse type of Lauren, andparticularly of the subgroup macroscopic classificationof Type III.

DISCUSSIONDISCUSSIONDISCUSSIONDISCUSSIONDISCUSSION

In patients with gastric cancer without evidenceof distant metastases, the main factor that determines theextent of the surgical procedure is the possibility of lymphnode metastasis1. These are also the main prognostic factorin EGC, especially when one considers that metastasis toother sites are very rare13. The incidence of these lymphaticimplants is greater the higher the degree of penetration ofthe tumor in the gastric wall. Thus, in advanced tumors,the rate of lymph node metastases ranges from about 50%,

Table 2 -Table 2 -Table 2 -Table 2 -Table 2 - Incidence of lymph node metastasis and N staging according to the degree of tumor penetration of the gastric

wall (t).

T4 (14)T4 (14)T4 (14)T4 (14)T4 (14) T3 (59)T3 (59)T3 (59)T3 (59)T3 (59) T2 (17)T2 (17)T2 (17)T2 (17)T2 (17) T1b (14)T1b (14)T1b (14)T1b (14)T1b (14) T1a (12)T1a (12)T1a (12)T1a (12)T1a (12)

N 0N 0N 0N 0N 0 7,1% 15,2% 41,2% 57,1% 83,4%

N 1N 1N 1N 1N 1 28,6% 11,9% 23,5% 14,3% 8,3%

N 2N 2N 2N 2N 2 7,1% 25,4% 23,5% 21,4% 8,3%

N3aN3aN3aN3aN3a 28,6% 32,2% 11,8% 7,1% 0%

N 3 bN 3 bN 3 bN 3 bN 3 b 28,6% 15,2% 0% 0% 0%

N +N +N +N +N + 92,9 %* 84,8 %* 58,8 %* 42,9 %* 16,7 %*

*p< 0.0001 (significant Association between the degree of tumor penetration of the gastric wall – T – and the presence of lymph nodemetastasis – N).

Table 3 -Table 3 -Table 3 -Table 3 -Table 3 - Tumor Characteristics and N staging of patients with EGC.

Pat ientPat ientPat ientPat ientPat ient T 1T 1T 1T 1T 1 Size(cm)Size(cm)Size(cm)Size(cm)Size(cm) Macro Classif.Macro Classif.Macro Classif.Macro Classif.Macro Classif. LaurenLaurenLaurenLaurenLauren Degree of DifferentiationDegree of DifferentiationDegree of DifferentiationDegree of DifferentiationDegree of Differentiation NNNNN

11111 b 1.6 III Intestinal Moderately Differentiated 0

22222 a 3.5 IIb Diffuse Poorly Differentiated 0

33333 a 0.5 I Intestinal Well Differentiated 0

44444 b 2.2 III Intestinal Well Differentiated 0

55555 b 1.8 IIa + IIc Intestinal Moderately Differentiated 0

66666 a 1.5 IIc + III Diffuse Poorly Differentiated 0

77777 b 1.5 III Diffuse Poorly Differentiated 0

88888 a 3.0 IIa Intestinal Well Differentiated 0

99999 a 1.7 I Intestinal Well Differentiated 0

1 01 01 01 01 0 b 1.5 IIa Diffuse Poorly Differentiated 0

1 11 11 11 11 1 b 2.2 III Diffuse Poorly Differentiated 0

1 21 21 21 21 2 a 3.0 IIb Diffuse Poorly Differentiated 0

1 31 31 31 31 3 a 3.0 IIa + IIb Diffuse Poorly Differentiated 0

1 41 41 41 41 4 b 3.5 IIb Intestinal Moderately Differentiated 0

1 51 51 51 51 5 a 2.0 III Intestinal Well Differentiated 0

1 61 61 61 61 6 b 1.8 III Intestinal Well Differentiated 0

1 71 71 71 71 7 a 4.3 IIa + III Intestinal Well Differentiated 0

1 81 81 81 81 8 a 2.0 IIa + III Intestinal Moderately Differentiated 0

1 91 91 91 91 9 b 2.2 III Intestinal Well Differentiated 2

2 02 02 02 02 0 b 0.8 IIc Diffuse Poorly Differentiated 2

2 12 12 12 12 1 a 3.5 III Diffuse Poorly Differentiated 2

2 22 22 22 22 2 a 3.5 III Diffuse Moderately Differentiated 1

2 32 32 32 32 3 b 0.4 III Intestinal Moderately Differentiated 2

2 42 42 42 42 4 b 1.5 III Diffuse Poorly Differentiated 3a

2 52 52 52 52 5 b 2.0 III Intestinal Moderately Differentiated 1

2 62 62 62 62 6 b 6.0 III + IIb Diffuse Poorly Differentiated 1

14

Rev. Col. Bras. Cir. 2014; 41(1): 011-017


when there is involvement of the muscularis propria (T2),to more than 80% when there is invasion of the serosa(T4). In these cases, surgery with curative potential isextended gastrectomy with D2 lymphadenectomy, as haslong been advocated by the Japanese School9,14.

In early tumors, ie, those in which there is invasiononly of the mucousa or at maximum of the submucosa,nodal metastases rates are lower, ranging from 0 to 20.3%(mean 3.2%) when the tumor is confined to the mucosaand from 10.2 to 33.3% (mean 19.2%) 2 when reachingthe submucosa. These variations are determined by thedegree of depth of invasion in each of these layers, withthe lowest rates of metastases in tumors with involvementof only the top 1/3 of the mucosa (m

1) and highest in those

with invasion of the lower 1/3 of the submucosa (sm3),

where the lymphatic network is richer15. Moreover, mostof these lymph node metastases in EGC reach the level 1,

perigastric lymph nodes, with an incidence of implants inlevel 2 chains ranging from 10% to 32%16,17. These datasuggest that extensive gastric resection with broad D2lymphadenectomy may be unnecessary for many patientswith EGC16. Smaller gastrectomies, with modified D1lymphadenectomy, have lower postoperative morbidity andmortality and provide better quality of life for this group ofpatients, without compromising the oncological radicalitynecessary for the treatment of gastric cancer18,19. Theresearch of sentinel lymph node in EGC can help determi-ne these nodal metastases20-23. Abe et al. proposed thecombination of EGC endoscopic resection with laparoscopiclymphadenectomy in patients with EGC with risk criteriafor lymph node metastasis24.

In our study, the number of patients diagnosedwith EGC, 26 (16.3%), was relatively high by Westernstandards, where this ratio is around 12-15%, but low

Table 4 -Table 4 -Table 4 -Table 4 -Table 4 - Lymph node metastasis and N staging according to T, size of the lesion, macrocospic classification, degree of tumoral

differentiation and histological type of Lauren.

TTTTT N +N +N +N +N + N 1N 1N 1N 1N 1 N 2N 2N 2N 2N 2 N 3N 3N 3N 3N 3

T1a 2/12 (16.7%) 1/12 (8.3%) 1/12 (8.3%) 0

T1b 6/14 (42.9%) 2/14 (14.3%) 3/14 (21.4%) 1/14 (7.1%)

Size of the lesionSize of the lesionSize of the lesionSize of the lesionSize of the lesion

< ou = 2cm 4/14 (28.6%) 1/14 (7.1%) 2/14 (14.3%) 1/14 (7.1%)

> 2cm 4/12 (33.3%) 2/12 (16.7%) 2/12 (16.7%) 0

Macrosc. Classif.Macrosc. Classif.Macrosc. Classif.Macrosc. Classif.Macrosc. Classif.

Types I and II 1/10 (10%) 0 1/10 (10%) 0

Type III 7/16 (43.8%) 3/16 (18.8%) 3/16 (18.8%) 1/16 (6.2%)

Degree of Diferent.Degree of Diferent.Degree of Diferent.Degree of Diferent.Degree of Diferent.

Well Dif. 1/8 (12.5%) 0 1/8 (2.5%) 0

Mod. /Poorly Dif. 7/18 (38.9%) 3/18 (16.7%) 3/18 (16.7%) 1/18 (5.5%)

Classif. of LaurenClassif. of LaurenClassif. of LaurenClassif. of LaurenClassif. of Lauren

Intestinal 3/14 (21.4%) 1/14 (7.1%) 2/14 (14.3%) 0

Diffuse 5/12 (41.7%) 2/12 (16.7%) 2/12 (16.7%) 1/12 (8.3%)

Table 5 -Table 5 -Table 5 -Table 5 -Table 5 - Relationship between rates of lymph node metastasis according to T, size of the lesion, macroscopic classification,

degree of tumor differentiation and histological type of Lauren.

T1aT1aT1aT1aT1a T 1 bT 1 bT 1 bT 1 bT 1 b T1 a<or= 2cmT1 a<or= 2cmT1 a<or= 2cmT1 a<or= 2cmT1 a<or= 2cm T1 a> 2cmT1 a> 2cmT1 a> 2cmT1 a> 2cmT1 a> 2cm T1 b<ou= 2cmT1 b<ou= 2cmT1 b<ou= 2cmT1 b<ou= 2cmT1 b<ou= 2cm T1b> 2cmT1b> 2cmT1b> 2cmT1b> 2cmT1b> 2cm

Size of the lesionSize of the lesionSize of the lesionSize of the lesionSize of the lesion

< or = 2cm 0 /50 /50 /50 /50 /5 4/9 - - - -

> 2cm 2/7 2/5 - - - -

Macrosc. Classif.Macrosc. Classif.Macrosc. Classif.Macrosc. Classif.Macrosc. Classif.

Types I / II 0 / 60 /60 /60 /60 /6 1/4 0 /20 /20 /20 /20 /2 0 /40 /40 /40 /40 /4 1/3 0/1

Type III 2/6 4/10 0 /30 /30 /30 /30 /3 2/3 3/6 2/4

Degree of Diferent.Degree of Diferent.Degree of Diferent.Degree of Diferent.Degree of Diferent.

Well Dif. 0 / 50 /50 /50 /50 /5 1/3 0 /30 /30 /30 /30 /3 0 /20 /20 /20 /20 /2 0/1 1/2

Mod. /Poorly Dif. 2/7 5/11 0 /20 /20 /20 /20 /2 2/5 4/8 1/3

Classif. of LaurenClassif. of LaurenClassif. of LaurenClassif. of LaurenClassif. of Lauren

Intestinal 0 / 60 /60 /60 /60 /6 3/8 0 /40 /40 /40 /40 /4 0 /20 /20 /20 /20 /2 2/5 1/3

Diffuse 2/6 3/6 0 /10 /10 /10 /10 /1 2/5 2/4 1/2


Rev. Col. Bras. Cir. 2014; 41(1): 011-017

compared to Eastern data2, 25. The overall incidence of lymphnode metastasis, 30.8% in our patients, was also very highwhen compared with both Eastern and Western patientpopulations. A German study with 126 patients with EGCshowed rates of lymph node metastases around 18%,ranging from 10.6% in restricted mucosal lesions to 25.3%in those with submucosal invasion15. Our rates of metastasisof tumors confined to the mucosa and with the invasion ofthe submucosa were 16.7% and 42.9%, respectively, alsohigh, and statistically significant difference between them.

However, the degree of tumor penetration of thegastric wall is just one of the factors determining the risk ofsecondary lymph node implants. Other important factorsare the macroscopic appearance of the lesion, tumor size,presence of lymphatic or vascular invasion and tumordifferentiation grade1,2,3,5-7,13,15. Early type III tumors, orulcerated, or mixed and partially ulcerated, display higherrates of metastasis than other types of EGC, as attested inour sample. A study of 684 patients with EGC showed alymph node metastasis rate of 3.4% in tumors confined tothe mucosa and 19% in those with invasion of thesubmucosa: in all patients with lymph node metastases thetumor had areas of ulceration16.

Tumor size is another risk factor for thedevelopment of lymph node metastases. The greater theinjury, the greater the risk of metastases1-16. In our series,the incidence of metastasis in tumors larger than 2cm wasgreater than in tumors smaller than 2cm. This parameter,although significant in univariate analyzes, is best evaluatedwhen considered in conjunction with other risk factors formetastasis. Hölscher et al., for example, demonstrated that

the probability of lymph node metastases is very low in T1atumors smaller than 2cm and in T1b lesions smaller than1cm15. Hirasawa et al.studied over 3800 patients and didnot observe lymph node metastasis in intramucosal, welldifferentiated, tumors of less than 3cm, withoutangiolymphatic invasion, whether ulcerated or not6.Regarding the degree of cell differentiation, well-differentiated tumors have a lower rate of metastasis thantheir moderately differentiated and poorly differentiatedcounterparts. Accordingly, as for Lauren’s classification, theintestinal type tends to have a lower rate of lymph nodemetastasis than the diffuse one26. These data are confirmedin our work and can be seen in Table 5. Still from thehistological point of view, tubular and papillaryadenocarcinomas have better prognosis than carcinomaswith signet ring cells13,15. Thus, it can be stated that theincidence of lymph node metastasis is null in tumors smallerthan 2cm, type I (elevated), well-differentiated, intestinaltype of Lauren and restricted to the mucosa, allowing thesafe conduct of endoscopic resection. On the other hand,tumors poorly differentiated or undifferentiated, diffuse typeof Lauren, larger than 3cm, even restricted to the mucosa,have rates of lymph node metastasis higher than 8%,reaching more than 20% when there is involvement of thesubmucosa 6. In our series, the incidence of nodalmetastases in EGC in these three conditions was 38.9%,41.7% and 33.3%, respectively.

Gotoda8 analyzed a large database of over 5000patients who underwent gastrectomy with meticulous D2lymphadenectomy, confidently establishing the risk of lymphnode metastases in EGC and listing four groups of patients

Table 6 -Table 6 -Table 6 -Table 6 -Table 6 - Relationship between rates of lymph node metastasis and injury variables: T, lesion size, macroscopic classification,

tumor differentiation grade and histological type of Lauren.

Var iableVar iableVar iableVar iableVar iable positive (n=8)positive (n=8)positive (n=8)positive (n=8)positive (n=8) negative (n=18)negative (n=18)negative (n=18)negative (n=18)negative (n=18) p valuep valuep valuep valuep valueaaaaa

nnnnn %%%%% nnnnn %%%%%

TTTTT

T1a 2 25.0 10 55.6 0.15

T1b 6 75 .075 .075 .075 .075 .0 8 44 .444 .444 .444 .444 .4

Size (cm)Size (cm)Size (cm)Size (cm)Size (cm)bbbbb 2.1 (0.4 – 6.0) 2.0 (0.5 – 4.3) 0.18 b

average = 2.5 cm average = 2.2cm

< ou = 2cm 4 50.0 10 55.6 0.56

> 2cm 4 50.0 8 44.4

Macrosc. Classif.

Types I e II 1 12.5 9 50.0 0.081

Type III 7 87 .587 .587 .587 .587 .5 9 50 .050 .050 .050 .050 .0

Degree of Difer.

WellDif. 1 12.5 7 38.9 0.45

Mod. /poorly Dif. 7 87 .587 .587 .587 .587 .5 11 61 .161 .161 .161 .161 .1

Lauren

Intestinal 3 37.5 11 61.1 0.24

Difuso 5 62 .562 .562 .562 .562 .5 7 38 .938 .938 .938 .938 .9

aaaaa Fisher exact test.bbbbb lesion size in cm was expressed as median (minimum - maximum) and compared by Mann-Whitney test.

16

Rev. Col. Bras. Cir. 2014; 41(1): 011-017


with a metastases rate of 0%: 1- tumor restricted tomucosa, well or moderately differentiated, withoutlymphatic vascular invasion, with or without ulceration,less than 3cm in size; 2- tumor restricted to the mucosa,well or moderately differentiated, without lymphaticvascular invasion, without ulceration, of any size; 3- tu-mor limited to the mucosa, undifferentiated or poorlydifferentiated, without lymphatic vascular invasion, withoutulceration, less than 2cm in size; and 4- tumor with su-perficial invasion of the submucosa (sm

1) well or moderately

differentiated, without lymphatic vascular invasion, lessthan 3cm in size.

Kwee et al., in a meta-analysis, sought to identifypredictive factors of nodal metastases in EGC2. The mainvariables significantly associated with metastases in tumorsrestricted to the mucosa were: age less than 57 years, tu-mor located in the middle 1/3 of the stomach, large lesions,depressed and ulcerated, undifferentiated, tumors, diffuseLauren type, and presence of lymphatic invasion. The mainvariables significantly associated with metastasis in tumorswith invasion of the submucosa were, on their turn, were:female gender, tumor located in the lower 1/3 of thestomach, large lesions, tumors with deeper submucosalinvasion, undifferentiated tumors, and lymphatic or vascularinvasion.

Other aspects related to lymph node metastasisstill not well understood and that may have negative impacton the survival of patients suffering from EGC undergoingresections and more conservative lymphadenectomy arethe micrometastases and the skip metastases. The formerare not detected in conventional histopathology, but onlywith immunohistochemical tests, and may therefore gounnoticed on frozen section exams, not being resected inD2 lymphadenectomy.

Kim et al. studied 90 patients staged as T1N0and found micrometastases in 10% of them27. None of

them presented with tumor recurrence in more thanfive years of follow-up. The main independent risk factorsassociated with these micrometastases were thepresence of lymphatic invasion and tumor size. The skipmetastases, those that occur on more distant lymphnodes chains without involvement of lymph nodesclosest to the tumor, bring the risk of false negativityfor metastases upon research for sentinel lymph nodes.Kitagawa et al. performed a study on sentinel lymphnodes and drew attention to a 5.1% rate of skipmetastasis28. Saito et al. analyzed 313 patients withlymph node metastasis in N2 chains and found 21(6.7%) without metastases in level 1 lymph nodes29.Most of these skip metastases concentrated on lymphnode chains 7 and 8, which should therefore be includedin the lymphadenectomy of patients with negativesentinel nodes undergoing minor gastric resection withmodified D1 lymphadenectomy10.

Our rate of 16.3% of EGC amongst the 160patients studied was relatively high for Westernstandards, but still very low when compared with Easternstandards. The incidence of lymph node metastasis inthese patients was very high and suggests that oneshould keep the radical ity of surgery for EGC,particularly in relation to D2 lymphadenectomy,recommended for advanced gastr ic cancer.Conservative resections with lymphadenectomies lowerthan D2 should be performed only in selected cases,well studied for the risk factors of lymph nodemetastasis. Despite the small number of cases did notallow to establish relation between the rate of lymphnode metastasis and the risk factors studied, there wasa strong tendency for the occurrence of these metastasesin tumors with invasion of the submucosa, larger than2cm, type III, poorly differentiated and diffuse type ofLauren, as already demonstrated in other works.

R E S U M OR E S U M OR E S U M OR E S U M OR E S U M O

ObjetivoObjetivoObjetivoObjetivoObjetivo: avaliar a incidência de metástases linfonodais no câncer gástrico precoce identificaando fatores de risco para o surgimento

destas metástases. MétodosMétodosMétodosMétodosMétodos: estudo prospectivo de pacientes portadores de câncer gástrico, internados na Seção de Cirurgia

Esôfago-Gástrica do Serviço de Cirurgia Geral do HUCFF-UFRJ, no período de janeiro de 2006 a maio de 2012. ResultadosResultadosResultadosResultadosResultados: a

frequência de câncer gástrico precoce foi 16,3%. A incidência de metástases ganglionares foi 30,8% e ocorreu com maior

frequência nos pacientes portadores de tumores com comprometimento da submucosa (42,9%), naqueles pouco diferenciados

(36,4%), nos tumores maiores que 2cm (33,3%) e nas lesões ulceradas do tipo III (43,8%). ConclusãoConclusãoConclusãoConclusãoConclusão: a incidência de metástases

linfonodais entre os pacientes foi muito alta e sugere que se deva manter, no câncer gástrico precoce, a radicalidade das ressecções,

particularmente no que se refere à linfadenectomia D2, preconizada para o câncer gástrico avançado. Ressecções conservadoras,

com linfadenectomias menores que D2 devem ser realizadas apenas em casos selecionados, bem estudados quanto aos fatores de

risco de metástases linfonodais. Apesar do pequeno número de casos não ter permitido relacionar o índice de metástases linfonodais

aos fatores de risco estudados, pôde-se verificar uma forte tendência à ocorrência destas metástases em tumores do tipo III, maiores

que 2cm, pouco diferenciados e do tipo difuso de Lauren.

DescritoresDescritoresDescritoresDescritoresDescritores: Neoplasias gástricas. Metástase linfática. Gastrectomia. Linfadenectomia. Fatores de risco.


Rev. Col. Bras. Cir. 2014; 41(1): 011-017

REFERENCESREFERENCESREFERENCESREFERENCESREFERENCES

1. Novotny AR, Schuhmacher C. Predicting lymph node metastasesin early gastric cancer: radical resection or organ-sparing therapy?Gastric Cancer. 2008;11(3):131-3.

2. Kwee RM, Kwee TC. Predicting lymph node status in early gastriccancer. Gastric Cancer. 2008;11(3):134-48.

3. Kunisaki C, Takahashi M, Nagahori Y, Fukushima T, Makino H,Takagawa R, et al. Risk factors for lymph node metastasis inhistologically poorly differentiated type early gastric cancer.Endoscopy. 2009;41(6):498-503.

4. Ishikawa S, Togashi A, Inoue M, Honda S, Nozawa F, Toyama E, etal. Indications for EMR/ESD in cases of early gastric cancer:relationship between histological type, depth of wall invasion, andlymph node metastasis. Gastric Cancer. 2007;10(1):35-8.

5. Li C, Kim S, Lai JF, Oh SJ, Hyung WJ, Choi WH, et al. Risk factors forlymph node metastasis in undifferentiated early gastric cancer.Ann Surg Oncol. 2008;15(3):764-9.

6. Hirasawa T, Gotoda T, Miyata S, Kato Y, Shimoda T, Taniguchi H,et al. Incidence of lymph node metastasis and the feasibility ofendoscopic resection for undifferentiated-type early gastric cancer.Gastric Cancer. 2009;12(3):148-52.

7. Jeong O, Ryu SY, Park YK. Accuracy of surgical diagnosis indetecting early gastric cancer and lymph node metastasis and itsrole in determining limited surgery. J Am Coll Surg. 2009;209(3):302-7.

8. Gotoda T. Endocopic resection of early gastric cancer. GastricCancer. 2007;10(1):1-11.

9. Japanese Gastric Cancer Association. Treatment guidelines forgastric cancer in Japan. 2nd ed. Tokyo: Kanehara; 2004.

10. Bravo Neto GP, Santos EG, Loja CAS, Carvalho FV, Neves MS,Pinto MF, et al. Ressecções gástricas menores com linfadenectomiamodificada em câncer gástrico precoce com linfonodo sentinelanegativo. Rev Col Bras Cir. 2012;39(3):183-8

11. Japanese Gastric Cancer Association. Japanese Classification ofGastric Carcinoma – 2nd English Edition. Gastric Cancer. 1998;1:10-24.

12. NCCN Clinical Practice Guidelines in Oncology (NCCN GuidelinesTM).Gastric Cancer (Including cancer in the proximal 5cm of thestomach). Version 2; 2011.

13. Ohashi S, Okamura S, Urano F, Maeda M. Clinicopathologicalvariables associated with lymph node metastasis in submucosalinvasive gastric cancer. Gastric Cancer. 2007;10(4):241-50.

14. Linhares E, Sano T, Kesley R. Linfadenectomia no câncer gástrico.In: Bravo Neto GP, Salles RARV, editores. Programa de Atualiza-ção em Cirurgia (PROACI), Ciclo 2, Módulo 1. Porto Alegre: Artmed/Panamericana; 2005. p.61-79.

15. Hölscher AH, Drebber U, Mönig SP, Schulte C, Vallböhmer D,Bollscweiler E. Early gastric cancer: lymph node metastasis startswith deep mucosal infiltration. Ann Surg. 2009;250(5):791-7.

16. Shimoyama S, Seto Y, Yasuda H, Mafune K, Kaminishi M. Concepts,rationale, and current outcomes of less invasive surgical strategiesfor early gastric cancer: data from a quarter-century of experiencein a single institution. World J Surg. 2005;29(1):58-65.

17. Ohdaira H, Nimura H, Mitsumori N, Takahashi N, Kashiwagi H,Yanaga K. Validity of modified gastrectomy combined with sentinelnode navigation surgery for early gastric cancer. Gastric Cancer.2007;10(2):117-22.

18. Yaguchi Y, Tsujimoto H, Kumano I, Takahata R, Matsumoto Y,Yoshida K, et al. Sentinel node navigation surgery attenuates thefunctional disorders in early gastric cancer. Oncol Rep.2012;27(3):643-9.

19. Ishikawa K, Arita T, Ninomiya S, Bandoh T, Shiraishi N, Kitano S.Outcome of segmental gastrectomy versus distal gastrectomy forearly gastric cancer. World J Surg. 2007;31(11):2204-7.

20. Dong LF, Wang LB, Shen JG, Xu CY. Sentinel lymph node biopsypredicts lymph node metastasis in early gastric cancer: aretrospective analysis. Dig Surg. 2012;29(2):124-9.

21. Kelder W, Nimura H, Takahashi N, Mitsumori N, van Dam GM,Yanaga K. Sentinel node mapping with indocyanine green (ICG)and infrared ray detection in early gastric cancer: an accuratemethod that enables a limited lymphadenecomy. Eur J Surg Oncol.2010;36(6):552-8.

22. Ichikura T, Sugasawa H, Sakamoto N, Yaguchi Y, Tsujimoto H,Ono S. Limited gastrectomy with dissection of sentinel node stationsfor early gastric cancer with negative sentinel node biopsy. AnnSurg. 2009;249(6):942-7.

23. Ishigami S, Natsugoe S, Uenosono Y, Yanagita S, Matsumoto M,Okumura H, et al. Usefulness of sentinel node biopsy in laparoscopicpartial gastrectomy for early gastric cancer.Hepatogastroenterology. 2007;54(79):2164-6.

24. Abe N, Takeuchi H, Ohki A, Yanagida O, Masaki T, Mori T, et al.Long-term outcomes of combination of endoscopic submucosaldissection and laparoscopic lymph node dissection withoutgastrectomy for early gastric cancer patients who have a potentialrisk of lymph node metastasis. Gastrointest Endosc. 2011;74(4):792-7.

25. Jacob CE, Gama-Rodrigues J, Bresciani CJC, Zilberstein B, CecconelloI. Câncer Gástrico Precoce. In: Bravo Neto GP, Salles RARV, edito-res. Programa de Atualização em Cirurgia (PROACI), Ciclo 2, Módulo4. Porto Alegre: Artmed/Panamericana; 2006. p. 77-154.

26. Ito H, Inoue H, Ikeda H, Onimaru M, Yoshida A, Hosoya T, et al.Clinicopathological characteristics and treatment strategies in earlygastric cancer: a retrospective cohort study. J Exp Clin Cancer Res.2011;30:117.

27. Kim JJ, Song KY, Hur H, Hur JI, Park SM, Park CH. Lymph nodemicrometastasis in node negative early gastric cancer. Eur J SurgOncol. 2009;35(4):409-14.

28. Kitagawa Y, Fujii H, Mukai M, Kubota T, Otani Y, Kitajima M.Radio-guided sentinel node detection for gastric cancer. Br J Surg.2002;89(5):604-8.

29. Saito H, Tsujitani S, Ikeguchi M. Clinical significance of skipmetastasis in patients with gastric cancer. Gastric Cancer.2007;10(2):87-91.

Received on 07/11/2012Accepted for publication 05/01/2013Conflict of interest: none.Source of funding: none.

How to cite this article:How to cite this article:How to cite this article:How to cite this article:How to cite this article:Bravo Neto GP, Santos EG, Victer FC, Carvalho CES. Lymph nodemetastasis in early gastric cancer. Rev Col Bras Cir. [periódico na Internet]2014;41(1). Disponível em URL: http://www.scielo.br/rcbc

Address for correspondence:Address for correspondence:Address for correspondence:Address for correspondence:Address for correspondence:Guilherme Pinto Neto BravoEmail: [email protected]

lymph node metastasis in early gastric cancer · lymph node metastasis in early gastric cancer. of...

Documents